Rainfed agriculture constitutes around 80% of the world’s agricultural land, achieving the lowest on-farm crop yields and greatest on-farm water losses. Much of this land is in developing countries, including sub-Saharan Africa (SSA), where hunger is chronic. The primary constraint of rainfed agriculture—frequently experienced in SSA—is water scarcity, heightened by the unpredictability of season onset, erratic rainfall, as well as the inability of farmers to provide adequate soil and crop management. Farmers react differently to constraints, making a variety of choices—including the timing of planting, type of land cultivation, fertilization, and scattered fields, among many others. Limited information is available on the combined effects of these strategies for improving crop yield and water use efficiency (WUE). An experiment was co-conducted with farmers over four consecutive rainy seasons (2014–2018) in Tanzania, to evaluate these strategies for single and joint effects in improving yield and WUE on rainfed pearl millet (Pennisetum glaucum (L.) R.Br.). The treatments used were flat cultivation both without and with microdosing, as well as tied ridging without and with microdose interaction, with different planting dates depending on farmers’ decisions. Results show that farmers react differently to the early, normal, or late onset of the rainy season, and cumulative rainfall during its onset, which affects their decisions regarding planting dates, yield, and WUE. Microdose fertilization increases both the yield and WUE of pearl millet significantly, with greater effects obtained using tied ridging compared to flat cultivation. For low-income smallholder farmers in a semi-arid agroclimate, using tied ridging with microdosing during early planting is an effective response to spatiotemporal rainfall variability and poor soils.